The effectiveness of insect-pollination test to evaluate the level of self-incompatibility and their genetic analysis in radish (Raphanus sativus L.).

We have tried an insect-pollination self-incompatibility (SI) test to strictly evaluate the level of SI as a model for the actual F(1) seed production field using radish as experimental material. Twelve inbred lines, homozygous for the S-alleles, were used in the artificial self-pollination and the insect-pollination SI test. There was a positive correlation ( r = 0.606) between the results by these two methods. Some lines showed a low level of SI in the insect-pollination test despite showing a high level of SI in the artificial self-pollination test. On the other hand, no lines showing a low level of SI in artificial self-pollination had a high level of SI in insect-pollination. These results show that the insect-pollination SI test can be considered to be a more reliable and stricter method than the artificial self-pollination test with respect to an evaluation of SI levels. We have raised and analyzed an F(2) population and F(3) lines derived from an F(1) cross between a line showing a high level of SI (R00-04) and one showing a low level. The rate of self-seed settings of the F(2) population showed a binomial distribution. There were 39 high-level SI plants to 15 low-level SI plants. This result and F(3) progeny tests suggested that the high level of SI which R00-04 showed is controlled by a dominant gene.